Surpassing homochiral assemblies in chiroptical activity through enantiomeric imbalance Authors: Xin Wen, Runjia Wang, Li Zhang, Minghua Liu Published: 2026 ABSTRACT AbstractAmplifying asymmetry via supramolecular self-assembly is essential for understanding homochirality in biology and developing functional chiroptical materials. While the “majority rules” dictates the handedness of heterochiral assemblies, their chiroptical signals are typically weaker than those of homochiral ones. Here, we demonstrate that controlled enantiomeric imbalance can enhance chiroptical activity beyond that of homochiral assemblies. Using dansyl-glutamic acid diamide enantiomers, we show that fine-tuning enantiomeric excess (ee) triggers structural reorganization, yielding maximum circular dichroism (CD) and circularly polarized luminescence (CPL) intensities exceeding those of pure enantiomer assemblies at specific ee values. Simulations reveal that the incorporation of the minority enantiomer constructively forms “chiral oligomers”, which are composed of different ratios of enantiomers and optimize local chromophore alignment via effective pitch angle to amplify the chiroptical activity. This strategy overcomes the conventional signal attenuation from racemic or near-racemic mixing, enabling chiral imbalance–driven asymmetric amplification and offering a rational design principle for high-performance chiroptical materials.Subject terms: Self-assembly, Supramolecular polymers, Polymers Introduction Homochirality remains a mysterious phenomenon in living organisms, and elucidating its origin is a fundamental and highly active area of research1,2. Yet, the key lies in understanding how a minute initial enantiomeric imbalance could be amplified to the homochiral state observed in nature3–5. A particularly powerful strategy for amplification of chiroptical signals is supramolecular assembly6–10. This approach leverages weak and non-covalent interactions to enable molecular recognition and self-sorting, which can dramatically amplify a tiny enantiomeric excess into macroscopic chiral materials11–13. In many helical supramolecular polymers, a slight chiral bias introduced into the building blocks is markedly amplified, resulting in a single-handed helix preference for the entire assembly14–18. This phenomenon is governed by the “majority rules” principle, where the minority enantiomer is forced to adopt the helicity favored by the majority7,19,20, typically resulting in a non-linear relationship between optical activity, normally indicated by circular dichroism (CD) signals, and enantiomeric excess (ee)21–25. However, while the “majority rules” reported in previous work describes that a slight excess of one enantiomer dictates the helical handedness of the entire system, resulting in a non-linear amplification of chiroptical signals, the signal's intensity typically reaches its maximum only at the enantiopure state19,20. Moreover, studies based on the “majority rules” have so far been largely confined to CD, which reflects ground-state chirality. The influence of ee modulation on circularly polarized luminescence (CPL), a probe of excited-state chirality26–30, remains poorly understood. Circularly polarized luminescence is a unique photophysical phenomenon and is essential for applications in 3D displays31–33, optical imaging34–37 and photoelectronic devices38–40. Crucially, the potential of heterochiral interactions (interactions between L- and D-isomers) to influence entire function should not be undervalued. Heterochiral interactions facilitate the co-assembly of opposite enantiomers41–45, leading to supramolecular architectures with unique functional properties46–51. Thus, the “majority rules” effect in the supramolecular assemblies has driven us to explore a novel asymmetric amplification of asymmetry for circularly polarized luminescence and to more thoroughly account for the role of heterochiral interactions. We previously explored how achiral cosolvents act as an external trigger to invert the helical handedness of a pure enantiomer assembly and modulate the excited-state chirality of homochiral assemblies52. Here, we focused on the intrinsic heterochiral interactions between L- and D-enantiomers, which revealed that a controlled enantiomeric imbalance can produce a chiroptical enhancement even exceeding that of homochiral assemblies. This finding challenges the established “majority-rules” paradigm, providing new insights into molecular packing and exciton coupling in heterochiral systems. Using dansyl-glutamic diamide derivatives (DNSG) as building blocks (Fig. 1a, b), we showed that the introduction of a small opposite enantiomer amplified the chiroptical signal beyond that of the enantiopure system (Fig. 1d). We initially observed the distinct assembly behaviors between the homochiral and heterochiral assemblies: the racemic mixture self-assembled into a stable yellow-green fluorescent organogel, whereas the enantiopure systems formed cyan-emitting precipitates. A series of experiments further demonstrated the unique advantage of heterochiral interactions. Remarkably, we unveiled a pronounced “chiroptical activity enhancement” phenomenon, in which the CPL intensity at a low ee value (≈60%) exceeds that of the enantiopure assembly, setting our work apart from previous works (Fig. 1c, d). Theoretical simulations further revealed that, although adding slight opposite configuration reduces the overall coupling strength, the formation of “chiral oligomers” alters the local arrangement and increases the effective pitch angle, thereby enhancing the overall luminescence dissymmetry factor. Our work not only illustrates effective strategies for programming supramolecular assembly through precise chirality interaction but also provides fundamental insights into the asymmetry amplification of CPL, paving the way for the rational design of adaptive and highly efficient chiroptical materials. Molecular structures and schematic depiction of surpassing homochiral assemblies in chiroptical activity through enantiomeric imbalance. a Molecular structures of the enantiomers L-DNSG and D-DNSG. b Schematic illustration of homochiral self-assembly of L-DNSG and D-DNSG. c Plots of chiroptical activity versus enantiomeric excess (ee) in previous work compared to this work. d Schematic illustration of enantiomeric imbalance-regulated assembly, contrasting the conventional “Majority Rules” in previous work and the “chiroptical enhancement” in this work. Results Enantiomeric and racemic assembly of DNSG monomers The enantiomeric L-DNSG and D-DNSG were synthesized (Supplementary Fig. 1), which formed suspensions in acetonitrile (MeCN) and exhibited excellent luminescent properties52. In contrast to the homochiral systems, the equimolar racemic mixture (L-DNSG: D-DNSG = 1:1) formed a stable organogel (Fig. 2a, inset). In addition, the enantiomeric suspensions and the racemic gel displayed distinct fluorescence emissions, with the former showing a maximum at approximately 473 nm and the latter emitting at around 506 nm (Fig. 2a). Photophysical analysis revealed that the emission from the enantiomeric assemblies originated primarily from a locally excited (LE) state of the dansyl group, whereas that of the racemic gel was attributed predominantly to a twisted intramolecular charge transfer (TICT) state53,54. Fluorescence lifetime measurements showed lifetimes of 22.4 ns for L-DNSG and 21.2 ns for D-DNSG assemblies, while the racemic assembly exhibited a reduced lifetime of 14.4 ns (Fig. 2b, c). The quantum yield (QY) of the racemic assembly was 27.5% (Fig. 2c), significantly lower than those of the enantiomeric assemblies (55.9% for L-DNSG and 55.0% for D-DNSG). These differences are consistent with the intrinsic photophysical characteristics of the TICT state, which typically shows lower QY and shorter lifetime compared to the LE state55,56 Notably, in dilute solution, both enantiomeric and racemic systems exhibited identical emission peaks at 519 nm, corresponding to the TICT state (Supplementary Fig. 2). Significantly, through controlling the assembly form (enantiomeric vs. racemic), we achieved switchable emission between the LE and TICT states in the chiral DNSG system. Fluorescence analysis and morphology of enantiomeric and racemic assemblies. a Fluorescence spectra of L-DNSG, rac-DNSG, and D-DNSG assemblies. Inset: Photographs of sample vials under UV irradiation. b Emission decay curves of L-DNSG, rac-DNSG, and D-DNSG assemblies. c Comparison of fluorescence lifetime and QY values. SEM images of L-DNSG (d), D-DNSG (e), and rac-DNSG assemblies (f). FM images of L-DNSG (g), D-DNSG (h), and rac-DNSG assemblies (i). The morphologies of the L-DNSG, D-DNSG, and rac-DNSG assemblies were characterized using scanning electron microscopy (SEM), fluorescence microscopy (FM), and atomic force microscopy (AFM). SEM images revealed that L-DNSG assembled into right-handed helical nanofibers, whereas the D-DNSG formed left-handed nanofibers, confirming the transmission of molecular chirality to the nanoscale architecture (Fig. 2d, e). In contrast, the rac-DNSG assembly adopted an achiral nanobelt morphology (Fig. 2f), reflecting effective chirality cancellation upon heterochiral mixing. FM images not only corroborated these structural differences, revealing chiral helices for L/D-DNSG and achiral nanobelts for rac-DNSG (Fig. 2g–i and Supplementary Fig. 3), but also provided direct visual evidence of the chirality-dependent emission behavior. AFM images demonstrated that the helices of L-DNSG had a width of approximately 692 nm and a height of about 207 nm, whereas the nanobelts of rac-DNSG exhibited a width of around 720 nm with a height of around 73 nm (Supplementary Fig. 4). The formation of achiral nanobelts in the racemic system, in stark contrast to the chiral helices fromed by the enantiopure systems, indicated a clear preference for heterochiral assembly. This was further corroborated by the distinct emission and phase behaviors between the two systems. Enantiomeric excess (ee)-Dependent Assembly of DNSG Monomers Following the observation that chiral composition (enantiopure vs. racemic) governs distinct supramolecular assembly behaviors, we further systematically investigated the assembly behavior of DNSG as a function of enantiomeric excess. By varying the L-DNSG/D-DNSG molar ratio across the entire compositional range (0–100% D-DNSG), a series of assemblies with controlled ee values was prepared. Notably, a stable gel phase was observed across a wide D‑DNSG ratio from 20% to 80%, corresponding to ee = +60% ~ −60% (Supplementary Fig. 5). Fluorescence spectral analysis revealed a pronounced ee-dependent shift in emission profiles (Fig. 3d). As the composition changed from pure L-DNSG to the racemate and then to pure D-DNSG, the emission maximum underwent a continuous red-shift from 473 nm to 506 nm, followed by a blue-shift back to 473 nm (Fig. 3a and Supplementary Fig. 6). Fluorescence and absorption analysis of ee-regulated assemblies at various concentrations. a Variation of maximum emission wavelength (λem) of L/D-DNSG co-assemblies at different ee values and concentrations. b, c UV-vis absorbance spectra (b) and maximum absorbance wavelength (λabs) variation (c) for L/D-DNSG co-assemblies at 11.00 mM. d Fluorescence spectra and a photograph under UV light of L/D-DNSG co-assemblies at various ee values. e Fluorescent color changes of L/D-DNSG co-assemblies at different ee values and concentrations. In the absorbance spectra of the assembled systems, a gradual red-shift of the maximum from 358 nm to 374 nm was observed as the composition varied from the L-DNSG to the racemate, followed by a corresponding blue-shift back to 358 nm upon progressing the D-DNSG (Fig. 3b). A detailed analysis of the absorbance maximum revealed abrupt shifts when the D-DNSG composition exceeded 20% or fell below 80% (Fig. 3c), a trend closely resembling the changes observed in the emission maxima (Fig. 3a). The resemblance between the absorption and fluorescence spectral shifts underscores that the enantiomeric composition directly governs the optical properties of the supramolecular assemblies. Notably, in dilute solution, both enantiomeric and racemic systems exhibited identical absorption bands centered at 340 nm (Supplementary Fig. 7), indicating that the observed spectral variations originate from the assembly process. Furthermore, at a low concentration of 0.3 mM in the assembled state, the L-DNSG assembly showed absorption maxima at 257 nm and 343 nm, whereas the rac-DNSG assembly exhibited peaks at 260 nm and 343 nm (Supplementary Fig. 8). The slight red-shift in the rac-DNSG spectrum suggests a more pronounced J-type aggregation character in the racemic system compared to the enantiopure one. As the system transitions from dilute solution to low-concentration aggregates and finally to high-concentration aggregates, spectral differences between enantiomeric and racemic systems progressively emerge and intensify. Motivated by this pronounced concentration dependence, we systematically investigated the assembly behavior across a range of concentrations and ee values. Both the emission and phase behavior exhibited a clear dual dependence on concentration and enantiomeric excess (Fig. 3a, e). At low concentrations, all assemblies adopted TICT-dominated emission and formed gels. At high concentrations, however, a divergence emerged: systems with high ee switched to LE-preferred emission, while those with low ee maintained TICT-dominated emission; correspondingly, their physical states (gel or suspension) varied systematically with ee values (Supplementary Fig. 9, 10). These results highlight the successful synergistic control of both photophysical and macroscopic properties through the coupled tuning of concentration and chiral composition. Amplification of asymmetry in chiroptical activity We employed CD and CPL spectroscopy to investigate how chiroptical signals vary with enantiomeric composition. The CPL spectra exhibited emission peaks at wavelengths consistent with the fluorescence profiles. Notably, the CPL intensity initially increased and then decreased as the ee value was reduced (Fig. 4a and Supplementary Fig. 11). The pure enantiomers showed mirror-image CPL signals, with a positive signal for L-DNSG and a negative one for D-DNSG. Interestingly, in enantiomeric mixtures, the CPL signals were significantly amplified at ee values around ±0.6, while it nearly vanished at ee = 0. A plot of the luminescence dissymmetry factor (glum) against enantiomeric composition revealed a clear nonlinear relationship (Fig. 4c, purple line). Although the CPL signal generally followed the “majority rules” effect by aligning with the handedness of the dominant enantiomer, an unexpected amplification was observed. Specifically, glum reached a positive maximum at 20% D-DNSG and a negative maximum at 80% D-DNSG, corresponding to ee values of ±0.6. The CPL signal of assemblies at ee = ±0.6 was significantly stronger than that at ee = ±1.0 (Fig. 4f), providing strong evidence that the CPL amplification originated from an ordered and cross-stacking arrangement of L-DNSG and D-DNSG at the molecular level. Chiroptical properties of ee-regulated assemblies at different concentrations. CPL spectra of L/D-DNSG co-assemblies at concentrations of 11.00 mM (a) and 1.10 mM (b). c Luminescence dissymmetry factors (glum) variation of L/D-DNSG co-assemblies at different concentrations. d CPL spectra of post-mixed L-DNSG and D-DNSG assemblies at various molar ratios. e Comparison of glum variations between post-mixed and co-assembled systems. f Schematic illustration of the enhanced CPL signal generated by the co-assemblies at ee = 0.6 compared to enantiopure assemblies (ee = 1.0). CD spectra of L/D-DNSG co-assemblies at concentrations of 11.00 mM (g) and 1.10 mM (h). i Absorption dissymmetry factors (gabs) variation of L/D-DNSG co-assemblies at different concentrations. In the CD spectra, the self-assembly of pure enantiomers resulted in mirror-image profiles with positive and negative Cotton effects for L- and D-enantiomers, respectively, originating from supramolecular excitonic coupling within the chiral assemblies (Fig. 4g and Supplementary Fig. 12). As the ee value varied from ±1 to ±0.8, the CD signal showed an enhancement and a noticeable red-shift. A further variation to ee = ±0.6 led to a slight reduction in the CD peak near 400 nm, accompanied by the emergence of a new inverse peak around 315 nm, with the two peaks crossing near 360 nm. The CD signal eventually disappeared at ee = 0. The signal enhancement at ee = ±0.8 and the subsequent peak splitting at ee = ±0.6 indicate a substantial reorganization of molecular packing and an enhancement of excitonic coupling. The absorbance dissymmetry factor (gabs) also exhibited a nonlinear dependence on ee (Fig. 4i, purple line). In contrast to the conventional “majority rules”, where helicity (CD intensity) in mixtures is lower than that in pure enantiomers, both the CD intensity and gabs were amplified in enantiomeric mixtures, signifying an efficient expression of the supramolecular helical superstructure with a higher degree of helicity at specific ee values. The consistent observation of signal amplification in both CPL (excited state) and CD (ground state) spectra at ee = ±0.6 rules out macroscopic phase separation between racemic and enantiopure domains. Instead, these results point to a coherent change in molecular packing within a co-assembled system, leading to this well-defined yet anomalous chiral amplification behavior. To further corroborate this hypothesis, we prepared post-mixing samples by physically mixing preformed L-DNSG and D-DNSG assemblies at different molar ratios. In these samples, the CPL intensity weakened monotonically as the ee value decreased (Fig. 4d), and the plot of glum against ee revealed a linear relationship (Fig. 4e). The CPL signal of assemblies at ee = ±0.6 prepared by the co-assembly method was significantly stronger than that obtained from the post-mixing method. To further distinguish co-assembly from possible phase-segregation scenarios, we designed and compared three experimental scenarios at ee=0.6 (L:D = 8:2), representing different assembly modes: a physical mixture of pre-formed homochiral L- and D-assemblies (Scenario 1), a mixture of pre-formed racemic and homochiral assemblies (Scenario 2), and a co-assembly prepared via pre-mixing (Scenario 3) (Supplementary Fig. 13). Fluorescence spectra demonstrated that Scenario 1 retains a predominant LE emission, confirming that molecules remain trapped within segregated homochiral domains where rotation is restricted. In contrast, Scenarios 2 and 3 both exhibited TICT-dominated emission, indicating that our proposed co-assembly model (Scenario 3) was distinct from the separated homochiral packing mode represented by Scenario 1 (Supplementary Fig. 14). XRD analysis also distinguished the structural nature of the co-assembled state from the post-mixed systems. Unlike Scenarios 1 and 2, which superimposed the diffraction peak of separated assemblies, respectively, the co-assembly exhibited a contraction of the primary d-spacing, shifting from 6.40 nm (ee = 1) to 6.13 nm (Supplementary Fig. 15). This contraction indicated that the minority D-enantiomers are molecularly intercalated into the L-enantiopure chain rather than being excluded. The chiroptical property provided compelling evidence for distinction of three scenarios (Supplementary Fig. 16). In Scenarios 1 and 2, the CPL intensities displayed an attenuation effect, yielding signals significantly lower than those of the enantiopure assembly. Specifically, the phase segregation would constrain the total signal to be lower than that of the homochiral assembly. In contrast, only the co-assembled sample in Scenario 3 exhibited a chiroptical enhancement, with the CPL intensity surpassing that of the homochiral assembly. The chiroptical amplification is unattainable in heterogeneous mixtures, thereby providing definitive proof of a molecularly intermixed heterochiral phase. Thus, the comparative analysis of the three scenarios provided consistent evidence that rules out phase segregation or simple signal superposition as the origin of the observed phenomenon (Supplementary Table 1). We further examined the chiroptical signal and the manifestation of the “majority rules” effect at low concentrations. At a concentration of 1.10 mM, the CPL intensity gradually decreased as the ee value decreased (Fig. 4b and Supplementary Fig. 17), and the plot of glum against enantiomeric composition exhibited a sigmoidal relationship (Fig. 4c, red line). Similarly, in the CD spectra at the same concentration, a progressive reduction in intensity was observed around 360 nm with decreasing ee values (Fig. 4h and Supplementary Fig. 18). The slight nonlinear dependence of gabs on enantiomeric composition further confirmed the presence of a “majority rules” effect (Fig. 4i, red line and Supplementary Fig. 19). These sigmoidal trends in both glum and gabs at low concentration align with the conventional “majority rules” effect, where the chiroptical signal of a mixture is always lower than that of the pure enantiomers. To determine the concentration threshold for CPL signal enhancement, we systematically measured the CPL signal as a function of ee values at multiple concentration levels (Supplementary Fig. 20–24). Notably, as the concentration increased above 2.75 mM, the glum profile transitioned to an abnormal sigmoidal curve featuring two maxima near ee values of ±0.6–0.8 (Fig. 4c, orange line). This transition likely arises from enhanced coupling between chromophores due to the concentration increase, which facilitates the amplification of chiroptical signals. The concentration-dependent switch from normal to anomalous majority-rules effect offers a new strategy for designing self-assembled nanosystems with tailored circularly polarized luminescence properties. To clarify the influence of concentration effect on chiroptical property, we analyzed the concentration-dependent fluorescence behavior of the enantiopure L/D-DNSG (Supplementary Fig. 25). At low-concentration conditions, the I506nm/I473nm ratio is high (>1.2), indicating that the TICT state is dominant. In this condition, the assemblies are loosely associated, allowing the dansyl groups to rotate. This conformational flexibility leads to relatively disordered packing and weak excitonic coupling. At high-concentration conditions, the I506nm/I473nm ratio drops sharply to below 0.8. This shift indicates a transition where the LE state becomes dominant as the supramolecular assembly becomes increasingly dense. In addition, the computational analysis indicates that the heterochiral co-assembly facilitates a larger and more optimized pitch angle compared to the homochiral packing. The synergy between restriction state and pitch angle might be the fundamental origin of the switch from conventional to anomalous behavior. Packing modes of chirality-dependent assemblies To elucidate the structural basis of the chirality-dependent assemblies, we systematically characterized the molecular packing of DNSG assemblies with varying ee values using X-ray diffraction (XRD), Fourier transform infrared (FTIR), and differential scanning calorimetry (DSC) measurements to evaluate the effect of optical purity on the molecular packing. XRD measurements first revealed distinct molecular arrangements dependent on optical purity (Fig. 5a and Supplementary Fig. 26). The powder X-ray scattering profile of the enantiomeric L-DNSG assembly displayed several diffraction peaks with a d-spacing ratio of approximately 1: 1/2: 1/3: 1/4, 1/5, 1/6, and 1/7, which was consistent with a bilayer packing structure. These sets of d-spacings of assemblies, except for the racemate, from 6.48 to 6.02 nm, are larger than the length of a single L-DNSG but smaller than twice, which was speculated to correspond to the length of the packing unit of the bilayer arrangement with interdigitated hydrophobic chains. And these d-spacings exhibited a slight shift in the small-angle range with decreasing ee values (Fig. 5d), suggesting a change in the packing mode in the assemblies. The assembly of racemate formed a bilayer structure, as evidenced by diffraction peaks corresponding to a d-spacing ratio of 1: 1/2: 1/3: 1/4. The d-spacing of the racemate gave a value of 3.47 nm, which was further confirmed by the wide-angle X-ray scattering (WAXS) measurement (Supplementary Fig. 27). The smaller d-spacing value observed in the racemate, compared to those of other assemblies, suggests that it likely formed a sloping bilayer structure with much interdigitation of the alkyl chains57. XRD, FT-IR and DSC analysis of ee-regulated assemblies. XRD patterns (a) and the corresponding d-spacing variations (d) of L/D-DNSG co-assemblies. FT-IR spectra (b) and the corresponding N-H stretching vibrations (e) of L/D-DNSG co-assemblies. DSC thermograms (c) and the corresponding melting points (f) of L/D-DNSG co-assemblies. FTIR spectroscopy was employed to probe intermolecular interactions in the self-assembled states across different ee values, which displayed four characteristic bands corresponding to N–H stretching, amide I, amide II, and asymmetric S=O stretching vibrations (Fig. 5b and Supplementary Fig. 28). A gradual redshift in the N–H stretching vibration was observed from 3302 cm⁻¹ (ee = 1) to 3301 cm⁻¹ (ee = 0.2) and finally to 3298 cm⁻¹ (ee = 0), indicating an enhancement of hydrogen bonding as enantiopurity decreased (Fig. 5e). While the amide I band remained unchanged at 1645 cm⁻¹, the amide II band region showed clear evolution. The intensity at 1558 cm⁻¹ enhanced and the intensity at 1543 cm⁻¹ weakened with decreasing the ee values (Supplementary Fig. 28), characteristic of the hydrogen-bonded and free amide II band, which further confirming the strengthening of the hydrogen bonding interaction. Although the S=O stretching vibration maintained its position at 1324 cm⁻¹, its intensity decreased, and the band broadened significantly in racemic assemblies. Collectively, these spectral changes demonstrate that the racemic assembly forms stronger hydrogen bond interactions compared to enantiopure systems58, providing a molecular-level explanation for the observed differences in assembly morphology and chiroptical properties. The thermal behavior of supramolecular assemblies with varying ee values was investigated using differential scanning calorimetry (DSC) and thermogravimetric analysis (TGA) study (Fig. 5c, f and Supplementary Fig. 29, 30). The TGA curves indicated that both enantiopure and racemic DNSG assemblies exhibited high thermal stability across a broad temperature range (Supplementary Fig. 30). In DSC thermograms, the enantiopure assemblies showed a single and sharp melting point (Tm) at around 113°C. In contrast, the racemic system exhibited a broader melting transition shifted to a lower temperature (Fig. 5c), which suggested a mutual diluent effect and the formation of a co-assembly mode between L- and D-enantiomers20,59. In addition, the melting point of mixed systems revealed a distinct U-shaped dependence on enantiomeric composition, with a minimum around 103°C at ee values of 0, ±0.2, and ±0.4 (Fig. 5f). The trend also indicated that the nature of the racemic species was a racemic compound60,61. Notably, at D-DNSG ratios of 20% and 80%, a melt-recrystallization process was observed, which suggested that the metastable assemblies formed at these specific compositions and underwent the rearrangement and recrystallization during the heating process62. To elucidate the aggregation mechanisms of enantiopure (L/D-DNSG) and racemic (rac-DNSG) supramolecular assemblies, variable-temperature UV-vis (VT-UV-vis) spectroscopy was performed. The VT-UV-vis spectra of L-DNSG and rac-DNSG assemblies exhibited distinct variations upon increasing temperature (Fig. 6a, d). For the L-DNSG assembly, the absorption band at 343 nm decreased in intensity as the temperature increased from 293 K to 319 K, but subsequently increased upon further heating to 353 K. The two-stage change may reflect the disruption of hydrogen bonding followed by the collapse of π–π stacking interactions. In contrast, the rac-DNSG assembly showed a gradual decrease in the intensity of the 343 nm absorption band with increasing temperature. Plotting absorbance as a function of temperature further highlighted the differences between the two systems (Fig. 6b, e). Notably, the racemic assembly exhibited pronounced hysteresis during thermal disassembly, indicative of stronger hydrogen-bonding or higher binding energy between heterochiral compositions that enhances the thermal stability of the racemic system. Temperature-dependent UV-vis analysis and computational simulation. Temperature-dependent absorption spectra of L-DNSG (a) and rac-DNSG assembly (d) upon heating from 293 K to 353 K. Absorbance at λ = 343 nm as a function of temperature of L-DNSG (b) and rac-DNSG assembly (e). DFT-optimized structures and binding energies of the homochiral dimer (c) and heterochiral dimer (f). Furthermore, the VT-UV-vis analysis is also possible to distinguish the post-mixed L- and D-assemblies from the co-assembled assemblies at ee = 0.6 (Supplementary Fig. 31). The co-assembled assembly shows a significant thermal lag in its disassembly transition compared to the post-mixed assembly. The disassembly profile of the post-mixed ee=0.6 sample exhibited rapid responsiveness to temperature, which was more closely aligned with that of the homochiral ee = 1.0 assembly. This observation indicates that the molecules in the post-mixing system remain trapped in their original homochiral domains (L-assembly and D-assembly), exhibiting the thermal instability of the enantiopure assembly. The mechanism of enhancement of chiroptical activity To elucidate the origin of the amplification of chiroptical activity, especially the CPL signal, theoretical calculations were carried out. We first evaluated the intermolecular binding energies, as they fundamentally govern supramolecular organization. Density functional theory (DFT) calculations revealed that the heterochiral dimer possesses a significantly higher binding energy (LD: −40.41 kcal/mol) than its homochiral counterpart (LL: −37.00 kcal/mol), as shown in Fig. 6c, f. The computed higher binding energy confirms that incorporating the opposite enantiomer thermodynamically stabilizes the system, thus favoring the heterochiral co-assembly mode, which is consistent with our experimental observations. Detailed structural analysis of the dimers further confirmed that the larger binding energy of the heterochiral dimer originates from its more compact geometry, which facilitates a strengthened hydrogen bonding and van der Waals contacts compared to the homochiral system (Supplementary Figs. 32, 33). Systematic energy calculations for trimers and tetramers also revealed that any sequence containing an intercalated D-monomer is significantly more stable than its homochiral counterparts, establishing a thermodynamic preference for heterochiral co-assembly over self-sorting (Supplementary Table 2). To gain deeper insights into the molecular organization at the specific ee value of 0.6, we performed systematic geometry optimizations on various pentamer sequences with an L:D ratio of 4:1. Computational analysis identified an energetic preference for various heterochiral pentamers (Fig. 7a). Among all possible sequences, the LLDLL configuration, where the minority D-enantiomer is centrally intercalated within the L-enantiopure chain, was identified as the global energy minimum (defined as 0 kJ/mol). In contrast, placing the D-monomer at the terminal positions or off-center resulted in significant energy penalties: DLLLL (22.49 kJ/mol), LDLLL (28.56 kJ/mol), LLLDL (65.22 kJ/mol), and LLLLD (10.26 kJ/mol). The substantial energy gap between the centrally inserted LLDLL motif and other configurations suggests that “LLDLL” sequence might be a representative model with thermodynamic preference. We further performed additional calculations to double the sequence of the model from a pentamer to a decamer, and the result indicates that the LLDLLLLDLL motif remains consistently more stable than other arrangements (supplementary Table 3). These calculations are based on isolated oligomers and therefore reflect short-range interactions. While they help rationalize the molecular basis of the structural changes observed experimentally, they are not direct models of extended supramolecular stacks. Computational simulation regarding the mechanism of CPL enhancement. a Energy profiles of pentamer conformations. b Exciton coupling energy matrix (kJ/mol) for LLLLL and LLDLL pentamers. (blue and green indicate higher and lower values). c Top view of the intermolecular angles within the LLLLL and LLDLL conformations. Generally, when enantiomers co-assemble within an aggregate, the chromophore organization tends to become more disordered, leading to a reduction in excitonic coupling and a corresponding decrease in glum (Fig. 7b). However, in this specific system, the luminescence dissymmetry factor can be described by formula (1): where Jn,n+s represents the excitonic coupling between chromophores separated by s units, and φ is the pitch angle between adjacent chromophores. Subsequent DFT calculations on the “LLLLL” and “LLDLL” pentamers yielded luminescence dissymmetry factors (glum) of 2.92 × 10⁻³ and 4.91 × 10⁻³, respectively63–65. The higher glum value for the “LLDLL” pentamer aligns with our experimental results, demonstrating the role of enantiomeric excess in enhancing chiroptical activity. This result reveals that although doping reduces the overall coupling strength, the formation of “chiral oligomers” alters the local arrangement and increases the effective pitch angle, thereby enhancing the overall glum (Fig. 7c). Based on the experimental data and oligomer calculations, we propose a plausible structural scenario where the minority enantiomer is molecularly intercalated into the homochiral matrix. This intercalation induces a local reorganization of the packing geometry, optimizing the excitonic coupling between chromophores and thereby enhancing the dissymmetry factor. This work offers a new insight for improving the dissymmetry factor in supramolecular assemblies. Discussion In summary, this work elucidates the pivotal role of enantiomeric excess (ee) in modulating chiroptical activity enhancement within supramolecular systems. By designing and studying chiral dansyl-glutamic diamide conjugates (L/D-DNSG), we demonstrated that heterochiral co-assembly leads to distinct structural and functional outcomes compared to homochiral systems—most notably the formation of a stable fluorescent organogel in the racemic state, in contrast to the precipitates formed by enantiopure components. Remarkably, we observed a non-monotonic enhancement effect in chiroptical activity, where the CPL signal was significantly amplified at low ee values and even surpassed that of the enantiopure assemblies. A series of experimental characterizations revealed that enhanced hydrogen bonding and ordered molecular packing were observed in heterochiral aggregates. These results suggest that the minority enantiomer serves as a structural modulator that thermodynamically stabilizes the heterochiral phase and potentially optimizes the excitonic coupling geometry. This study not only offers insights into asymmetry amplification of circularly polarized luminescence but also advances the design of chiroptical materials. Methods Materials All reagents and solvents were purchased from commercial suppliers and used without further purification. N-(tert-Butoxycarbonyl)-L-glutamic Acid (Boc-L-Glu) and N-(tert-Butoxycarbonyl)-D-glutamic Acid (Boc-D-Glu) were purchased from TCI. 1-Octadecylamine was brought from Macklin. Sodium Bicarbonate (NaHCO3) was brought from Alfa Aesar. 1-Hydroxybenzotriazole (HOBt) was purchased from J&K. Trifluoroacetic acid was purchased from Energy Chemical. 1-(3-Dimethylaminopropyl)−3-ethylcarbodiimide hydrochloride (EDC·HCl), Triethylamine, and Dansyl chloride (DNS) were purchased from Innochem. Trichloromethane (TCM), ethanol (EtOH), and acetonitrile (MeCN) were supplied by Concord Technology (Tianjin) Limited Liability Company. All aqueous solutions were prepared using Milli-Q water (18.2 MΩ·cm) in this investigation. Supramolecular assembly protocol Typically, L-DNSG or D-DNSG solid (11.00 mM) and 1 mL MeCN were added into a seal-capped vial. The sample was then heated until a transparent solution formed. Subsequently, the sample was sustained at room temperature, resulting in the formation of either a suspension or gel. The concentrations of DNSG were varied at 1.10, 2.75, 5.50, 8.25 and 11.00 mM, respectively. The enantiomeric excess values of DNSG were changed by altering the ratio of L-DNSG and D-DNSG while maintaining the overall concentration consistent. NMR spectra 1H NMR spectra were recorded on a Bruker Fourier 300 spectrometer (for 300 M). Mass spectra MALDI-TOF MS mass spectra were recorded on a BIFLEIII matrix-assisted laser desorption/ionization time-of-fight mass spectrometry instrument. Ultraviolet-visible (UV-vis) spectroscopy UV–Vis spectra were recorded on a UV-2600 (Shimadzu) spectrophotometer or JASCO J-815 spectrometer. Samples were prepared for testing in quartz cuvettes with a 1 mm light path. All measurements were conducted in solvents of spectroscopic grade from commercial sources. The solid samples were dispersed into BaSO4 powders to record the UV-vis diffuse reflectance spectra on a UV-2600 (Shimadzu) spectrophotometer. Temperature-dependent UV-vis spectroscopy (heating) The solid powder of L-DNSG or rac-DNSG were weighed into quartz cuvettes with a light path length of 10 mm, and 2 mL MeCN was added to give the final measurement concentration to 1.10 mM. Then the samples were heated until dissolving clearly and cooled to room temperature. Afterwards, the measurement was conducted with a heating rate of 2 K/min without stirring and using a data interval of 2 K to minimize equilibration periods during the spectra acquisition. Fluorescence (FL) spectroscopy All fluorescence spectra were recorded on a F-4600 fluorescence spectrophotometer. For the all samples, the excitation (EX) slit and emission (EM) slit widths were set as 5 nm, and the PMT voltage was set as 400 V. Quartz cuvettes with an optical length of 1 mm were used. Circular dichroism (CD) spectroscopy CD spectra were recorded using JASCO J-815 and J-1500 spectrophotometers. Samples were prepared in quartz cuvettes with a light path length of 0.1 mm. Circularly polarized luminescence (CPL) spectroscopy CPL spectra were obtained using JASCO CPL-300 spectrophotometers. To keep the detector at the best state and obtain the precise signals, the DC values were monitored to approximately 0.5 V. The sample was prepared by transferring a droplet of suspension into a quartz cuvette with a light path length of 0.1 mm. The CPL spectra were tested at least 5 times to ensure the accuracy of the measurements. The glum spectra were transferred from CPL spectra using the Spectra Manager software of JASCO. Scanning electron microscopy (SEM) SEM images were recorded on a Hitachi S-4800 FE-SEM instrument with an accelerating voltage of 10 kV. The samples were prepared on single-crystal silicon wafers and dried under vacuum. Before SEM measurement, the samples on silicon wafers were coated with a thin layer of platinum (Pt) to enhance contrast. Fourier transform infrared (FT-IR) spectra FT-IR spectra were recorded on a BRUKER VERTEX 70 v spectrometer. The solid samples were prepared on KBr pellets after vacuum-drying. X-ray diffraction (XRD) measurements XRD patterns were obtained on a Rigaku D/Max-2500 X-ray diffractometer (Japan) with Cu Kα radiation (λ = 1.5406 Å), which were operated at a voltage of 45 kV and a current of 200 mA. Samples were cast on silicon substrates and dried under vacuum for XRD measurements. Quantum yield (QY) measurements The absolute fluorescence quantum yield was measured using a FluoroMax+ (HORIBA) spectrometer with a calibrated integrating sphere. Fluorescence lifetime measurements The delayed emission spectra were recorded on an Edinburg FLS-980 fluorescence spectrometer with an external 360 nm laser as the excitation light source using time-correlated single photon counting (TCSPC). Fluorescence microscope image (FM) The samples were observed on an IX 83 (Olympus) fluorescence microscope using high-pressure mercury lamp as excitation source for fluorescent images. The samples were prepared by dropping the assemblies with MeCN solvents onto the glass slide and naturally dried solvent. Atomic force microscope (AFM) The AFM images were recorded on a Dimension FastScan (Bruker Nano) instrument by using Fast-scan Mode with a Fast-scan B tip. The sufficiently dilute samples were cast onto single-crystal silica plates, the solvent was evaporated under the ambient conditions and dried under strong vacuum. Wide-angle X-ray scattering (WAXS) measurements The WAXS partterns were recorded on a Xeuss 2.0 equipment (Xenocs, France) with a sample-to-detector distance of 134 mm and an X-ray radiation wavelength of 1.5418 Å. Scattering patterns were collected by a semiconductor detector (Pilatus 300 K, DECTRIS, Swiss) with a resolution of 487 × 619 pixels (pixel size = 172 μm2). The samples were measured on silicon substrates, and each WAXS pattern was collected for 30 min. The final data were background-calibrated and fitted using SasView software. 1-D intensity curves could be generated by integrating the background-calibrated 2D-GIWAXS patterns. Differential scanning calorimetry (DSC) measurements The DSC thermograms were performed in an instruments DSC Q2000. Thermogravimetry analysis (TGA) The TGA was carried out using a HITACHI STA300 thermogravimetric analyzer. The composites were heated under nitrogen purge from 30 °C to 600 °C at 10 K min−1. Supplementary information Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Supplementary Data 4 Supplementary Data 5 Supplementary Data 6 Supplementary Data 7 Supplementary Data 8 Transparent Peer Review file Source data Source Data Acknowledgements We thank Dr. Meirong Liu from the Spectroscopy Group at the Institute of Chemistry, Chinese Academy of Sciences (ICCAS) for her help of spectroscopy measurements. Author contributions M. L. and L. Z. conceived the idea, designed the starting compounds and supervised the project. X. W., L. Z., and M. L. designed the experiments. X. W. carried out the experiments. R. W. made the computational simulation. X. W. and L. Z. analyzed the data. X. W., L. Z., and M. L. co-wrote the manuscript. Peer review Peer review information Nature Communications thanks the anonymous reviewers for their contribution to the peer review of this work. A peer review file is available. Funding This work was supported by the National Key R&D Program of China (2022YFA1204402), National Natural Science Foundation of China (52321006 and 92356302), Strategic Priority Research Program of the Chinese Academy of Sciences (XDB0770101 and XDB1300301) and the Chinese Academy of Sciences Project for Young Scientists in Basic Research (YSBR-027). Data availability All relevant data generated during and/or analyzed during the current study are available from the corresponding author. The spectroscopic, chiroptical, and calculation data generated in this study are provided in the Supplementary Information/Source Data file. Supplementary Information is available in the online version of the paper. Source data are provided with this paper. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Li Zhang, Email: zhangli@iccas.ac.cn. Minghua Liu, Email: liumh@iccas.ac.cn. Supplementary information The online version contains supplementary material available at 10.1038/s41467-026-74210-w. References Associated Data Supplementary Materials Supplementary Information Description of Additional Supplementary Files Supplementary Data 1 Supplementary Data 2 Supplementary Data 3 Supplementary Data 4 Supplementary Data 5 Supplementary Data 6 Supplementary Data 7 Supplementary Data 8 Transparent Peer Review file Source Data Data Availability Statement All relevant data generated during and/or analyzed during the current study are available from the corresponding author. The spectroscopic, chiroptical, and calculation data generated in this study are provided in the Supplementary Information/Source Data file. Supplementary Information is available in the online version of the paper. Source data are provided with this paper.